Summary
Jingu Kang is a Principal Software Engineer based in Edinburgh with 12 years of deep compiler and performance engineering experience, now driving compiler work at Arm. He specializes in ML compiler development (including GLOW) and competitive performance tuning across major benchmarks, combining LLVM/Clang expertise with kernel and codegen optimization for CPUs, DSPs and ML accelerators. His background spans C/C++/OpenCL compiler toolchains, contributions to projects like Google clspv and LLVM, and earlier GCC and Dragonegg work adding ARM ABI support. Jingu’s career began in firmware and device driver development, giving him a rare full-stack perspective from low-level firmware to high-level ML compiler optimizations. He is notable for turning benchmark-driven insights into pragmatic compiler optimizations such as vectorization, if-conversion and instruction scheduling. Colleagues rely on him for both deep technical fixes and cross-platform performance wins.
12 years of coding experience
20 years of employment as a software developer